[amishx64]

I have been planning my second build for nearly a year now. I have decided to build a second machine after completing my first Can-Not-Cut machine a few years ago. I have been picking up parts whenever I found deals on eBay and planning everything out. I have been designing and modeling virtually every part in SolidWorks for the past year as well and I think I have a good idea of how everything is going to go together. I would like criticisms and suggestions, however as I am sure there are things that could be improved. This is going to be a large (to me) router-style build.


Specs for my machine, dubbed ScratchPad:

Cutting Area:
~ 3’x3’x8”

Capabilities:
Wood, aluminum and some light steel.

Precision Goal:
X&Y: 200 steps/1” pitch = 200 --> .005”
Z: 200 steps/.5 pitch = 400 --> .0025”

Ballscrews:
2x Lamb Technicon (made ~1990) ground with 1 inch pitch double start (L105700157)
1x Thompson Ballscrew Assembly preloaded and rolled 0.5 inch lead (BSA-08-Q)

Electronics:
X: Pacific scientific PowerPac K43 with Gecko G203V @ 70V
Y: Pacific scientific Sigmax E42 with Gecko G203V @70V
Z: Chinese motor with Centent CN0142 @50V

Linear Rails:
X: 4x 500mm Hiwin LGR25TH Rails with LG25 bearings
Y: 2x 1500mm IKO LWH25 rails with LWHT25 Blocks
Z: 2x 500mm IKO LWH25 rails with LWHT25 Blocks

Frame:
2”x2”x.25” A500 hot square tube

Shaft Couplings:
X&Y: 2 sets of Boston FC20 Flexible Shaft Couplings
Z: Rigid .25in shaft coupling (Enco # 891-6307)

Spindle:
Hitachi M12VC and another yet to be determined

[amishx64]

Like I said, I have but a lot of work into my SolidWorks models, so I intend to show them off in case anyone finds any issues.
I have spent a ridiculous amount of time on is the Z axis. This assembly has gone through probably 9 overhaul revisions and hundreds of changes and there are still more things to change that I have yet to do, however it is mostly complete.

Specs:

Router plate:
11” wide x 350mm x 1/2” thick
Bolts on top of ‘Bridge’ plate. I will secure any router/spindle mounts to this plate. This has built-in setscrews to tram other mounted plates. I still have to add mounting holes.
I am also debating if this should be thicker. Input?

‘Bridge’ plate:
11” wide x 110mm x 1/2” thick
The ‘bridge’ connects the two bearing blocks and ballscrew nut. The ballscrew nut will bolt directly to the plate.

Base plate:
500mm x 270mm x 1in thick
There are two 9.97mm deep pockets for the ballscrew pillow blocks to sit snugly in.

14x M6 tapped holes for bolting 2x 500mm IKO LWH25 rails with LWHT25 Blocks
16x M8 holes for securing 4 LWHT25 Blocks for the X axis
8x 8-32 tapped holes for the Thompson ballscrew
10x M8 holes for securing the Lamb Technicon ballscrew
All these holes make the plate look like Swiss cheese.

Nema23 mount:
2.22”x2.22”x1.5” aluminum
This mount has holes so the stepper motor can bolt all the way through to the ballscrew pillow block.

The router plate is roughly two inches away from the base plate. This isn’t as close as I’d like, but it looks like it is about the best I can do with what I have. Shorter bearing blocks and a lower profile ballscrew assembly would have been ideal.

Since the M8 sockets are too tall (no clearance between top and bottom of IKO bearing block) and I couldn’t find any button head sockets that were 40mm long, I will have to grind down the tops of the bolts. Collision distance looks to be 1.2mm, so hopefully removing that much off the top of the bolt won’t cause problems down the road.

[amishx64]

The gantry the most lackluster parts of the machine. I am not a Mechanical or Structural Engineer so I am sure this is pretty bad structurally.

In my design the long length of steel holds the Y axis and the shorter sides will bolt to the X axis bearing blocks.

The gantry will measure 1327mm long x 900mm high x 100mm deep. Space between X axis rails will be 360mm. The material is 2”x2”x1/4” A500 hot steel. I will be welding it together with a MIG or Stick welder. I am planning on keeping things nice and square with the aid of two welding clamps and machinist squares. To level the surface to mount the rails, I will be using 105 epoxy resin and 206 hardener to apply an epoxy layer between 1/16” and 1/4” thick.

Any suggestions on a better gantry would be greatly appreciated. Note: SW shows some of the weldments as separate cuts. I will make my pieces as continuous as possible.

[steve123]

I want you to be able to cut steel with that, but you aren't there yet!

The gantry was the first thing that could use some strengthening. If you put a skin on the top, back, and bottom of the beam, and down the sides, it will turn it into more of a big single beam, that will be far stronger. You could also do it with diagonal braces (think triangles everywhere). I don't see you cutting steel unless you can get it to deflect less than .001" with 25 lbs of force. Also think about sideways force on the gantry. Will you be able to put force sideways on that beam, and keep it from parallelograming?

For your gantry design, the table down the x is not supported. It's taking it's own weight plus the gantry. It will sag. I think you should use some taller beams there.

To cut steel you would also need a spindle that can go under 3000 RPM and still have lots of torque (never a router). That usually means a belt-driven spindle. I noticed that you mentioned having a 2nd spindle.

The Y axis bearings look great.

The other thing I saw was that you have only two bearing blocks on the z axis. Bearing blocks aren't actually made to handle this kind of torque (even good ones). The force from the long lever arm from the bearing block to the end of the tool (about 8-10 inches?) divided by the short length of the bearing block will end up putting a very large force on the block, and you will get quite a bit of flex. You should have another set of bearing blocks below those, and they shouldn't be touching (at least a few inches between). This will make your Z back plate taller, and require longer linear rails, or you can decrease your z height. The other thing about z's is the plates themselves will flex, so it is a good idea to add some kind of I-beams to them instead of trying to just make them thicker.

The thing I see around here the most is people having to go back and make a stronger gantry beam, and fix a floppy Z.

For your precision goal. Those gecko drives will do 10x microsetpping, so you will actually do better than you think there. Your real precision problem will be gantry + z stiffness. The tool will go wherever it wants if things aren't stiff enough.

I am working on a design also, 3'x4'x6", made for cutting steel. I can post some pix if you are interested. I'm getting ready to start my build log.

I know you have done a lot of work so far! Keep at it!

-Steve

[amishx64]

Thanks for the great response Steve!

Your are correct about the gantry. Given enough side force, it would crumple into a sheet. Not sure why I didn't see that.
I have not bought any steel from my supplier yet, so I have an option to buy steel I-beams as well. I wanted to stay away from I-beams because I figured they would have more variation in the material, but that can change. Would you use I-beams over 2x2 square?

I will work on changes to the gantry and see what I can do with the Z axis. I'll see what my options are concerning rails & bearing blocks. I don't have access to aluminum C-channel or I-beam style aluminum, so maybe I can mock up a design with bolted on side lips that looks similar to C-channel. What plate would this be necessary on? The router plate (1/2" thick) or base plate (1" thick) or both?

The Y axis bearings look great.

Thanks. Guess I got something right.

I would love to see your build log when you get a chance to put it up. Keep me posted.

[louieatienza]

I would agree with Steve's observations about the gantry. A combination of plate and tubes would help greatly. If I'm not mistaken, the lower gantry rail will take on more force than the upper.

Also, I agree that the single Z bearings are not enough especially for steel. I might consider mounting the bearings on the Y plate and mounting the rails to the carriage plate.

[amishx64]

I would agree with Steve's observations about the gantry. A combination of plate and tubes would help greatly. If I'm not mistaken, the lower gantry rail will take on more force than the upper.

Noted. Should I reinforce the lower rail? I am not sure about using sheet. I would like to, but I don't have a plasma cutter or anything to cut plate.

I might consider mounting the bearings on the Y plate and mounting the rails to the carriage plate.

I am thinking the same thing. I'll do some work in SW and see what makes the most sense.

[amishx64]

Before I address all the adjustments needed to the Z axis and gantry, I wanted to post the info on the ballscrew assemblies.

Designing an assembly to house my Lamb Technicon ballscrews was pretty difficult given no technical data. Luckily, I was able to track down the company (MAG) that has the data regarding the ballscrews.

So far I have had a bushing machined to match the ballscrew and 7212 AC Bearings via an interference fit. The 7212 bearings were not what the Lamb technical diagram recommended (7314) but the 7314’s were much more expensive and larger than what I wanted to I went with the 7212’s.

Specs:

Ballscrew:
Lamb Technicon L105700157
2.250” thread diameter
1.00” pitch double start thread
Very heavy (85+ lbs?)

Aluminum Shaft Coupling:
ID Bore: 34.874mm
Thickness: 1/4”
1” shaft with 1/4” keyway
6x M8 Threaded Holes

FC-20 Coupling:
1” ID with keyway and other matching side with rubber spider
Connects directly to Pacific Scientific steppers

Bushing:
41.275mm ID
60mm small OD
79.5mm large OD
AC bearings are mounted face-to-face and held via an interference fit. To do this I baked the angular contact bearings at 325F and chilled (0F) the steel bushings and slid them together while still hot/cold.

7212 AC Pillow Block:
Material: Aluminum
Two halves are bolted together with M6 bolts.
Wall thickness: 22mm
I am still looking at making this thicker

Small Pillow Block:
2x 1614 Shielded Bearings
3/8” bolt connecting bearings and ballscrew

[amishx64]

I have revised the Z axis to include two more bearing blocks and longer rails for increased travel to compensate. I figure it is best to spend the money so that the ballscrew is fully utilized. I don’t know if I calculated the Z travel incorrectly last time, but it now looks like I will have around 10” of travel.

There is roughly 2” between the bearing blocks.

Specs:

Router plate:
11” wide x 475mm x 1/2” thick (125mm longer)
The router plate bolts on top of ‘Bridge’ plate and two other arbitrary plates connecting the back bearing blocks.
I am still debating if this should be thicker.

‘Bridge’ plate:
11” wide x 110mm x 1/2” thick
Specs have not changed.

Base plate:
500mm x 270mm x 1in thick (70mm longer)
500mm rails with two pillow blocks each.
No other changes aside from positioning differences and two more bolts for longer rails.

Side ‘Name’ Plates:
Two 1/4” thick aluminum plates secured with M6 bolts will provide C-channel-like structural support. These could be thicker.

Nema23 mount:
Specs have not changed.

Please let me know if you guys thing anything else needs changed.

[amishx64]

Well I changed the gantry and now it looks really heavy. I’m not really sure if this design is practical at all. I might have to go with putting in steel plate where some of the cross sections are due to weight and my inability to cut these pieces. I need to do some research on cutting steel plate.

The SolidWorks mass properties dialog using generic steel alloy as the material type lists the gantry at ~342lbs. Yikes!

The only other note is that the gantry is 70mm taller due to the Z axis change.

[painter]

Hello sir,
Thanks for posting on my build.
I think your new aproach to z axis is a smart choice. There is a lot of steel in your gantry redesign, I think that its to tall (to much leverage,/ distance frome your bearing trucks for cutting steel)
Take another look at my machine it has better potential to cut steel than the under table screw type machines, (Not my design I'm borrowing from others)!!!!
I have a bridgeport mill weighs 2100 lbs and often the table chatters when cutting steel and even aluminum! The lifting forces are incredidible when cutting steel. I hope to cut aluminum at times with my machine, so I have the rails above the cutting surface to help hold down the lift generated by the cutting bit /end mill and my bearing trucks are spaced as far apart as possible 14" !
I am using 3/16 wall 2x4 tubing in my design and I know it will not effectively cut steel , especialy when I see my bridgeport mill rumble and shake at times.
God bless and Merry Christmas to you, Tom

[amishx64]

Tom,

I see you have returned the favor. Thanks for posting on my build too!

The height of the machine is indeed a great worry to me as far as rigidity and flexing is concerned. I don't intend on using this machine for cutting mostly steel, but rather just very occasionally at very slow speeds. Thankfully steel is not my main focus.

I agree that your machine is much better setup to cut steel. You have a very short machine with drive screws on each side, making for a very rigid and flex-resistant design. I thought of doing a design similar to that when I was in the planning stages but I ultimately decided that I wanted a much higher Z clearance, leading me to my current router-based design. My main focus for this machine is wood, plastic, and aluminum and a very large area to machine these materials. I think I am on the right track to do this.

I really do appreciate the advice. Sadly there are not too many ways to reduce the distance from the ballscrew to the rails to the spindle head without drastically reducing Z travel. I could put the rails closer to the Z axis, but that would make them farther from the ballscrew creating off-center moment loading, which is a very big no-no. I think I am going to stick with the current layout for this purpose unless I can think of something really wonderful between now and build time.

Again, I appreciate the insight and I will try to minimize the height as much as possible. If you can find any other area that could be improved, please let me know. Best of luck on your machine and Merry Christmas all the same.

[painter]

Hello again,
Could you do something like this with your bearing trucks at the bottom of your gantry?
I realize this would shorten y travel however the gain against racking forces would be significant. With your open ended design perhaps you could reposition the workpiece your cutting to make up for travel loss of y, kind of like tiling, just thinking.
As to cutting plate (1/8 3/16 ) you could use a jigsaw/sabre saw with 10-14 bi-metal blades not the fastest but it works!

Tom

[painter]

Hello AGENNN,
Sorry can't seem to keep my opinion to myself today!
I think you could eliminate a lot of that steel in your gantry if you skinned it , .045 or .065 aluminum fastened at close intervals would make it surprisingly strong , would'nt be so top heavy. I used .045 alluminum in our racecar cockpit/tub and when all riveted together it became surprisingly strong... cuts pretty nice with a jigsaw/ sabre.
After coming back to my machine after a 5 year break from it I regret some cost saving decisions I made .... when it's all said and done there will be a buttload of money tied up in it and a couple hundred more would'nt be noticed in the end!

Merry Christmas
Tom

[painter]

waiting for my shop to heat up,
Had an image of a 300lb man with his ankles tied together... at 5'5'' 180lbs I think I could tip him over... however if his feet are untied and spread out there is no way I could tip him!

I hope my posts do not offend you.... I mean only the best for you at this early design stage of your soon upcoming build!!!
I wish I would have had input early on in my build, to late now tho, the junks welded together, believe me there are changes I would have made !!!!

Perhaps on my next build.... YIKES....

Sincerely Tom

[ger21]

Do you plan on mounting the spindle clamp to the bottom of the Z plate?
If so, I see two potential issues.
One, you're still going to get some flex, as the spindle will still be mounted to the end of a lever. It's just a shorter lever.
And two, if your spindle can reach the bottom of your table, I doubt it will clear the bottom of the gantry when fully retracted.

I'd recommend modelling the spindle and clamp, and seeing where different length tools end up at both the top and bottom of travel. You may either be able to shorten the Z plate, or spread the bearings further apart, both of which will make the Z axis stiffer.

[amishx64]

Tom,
Thanks for all the posts and insight. I will try a design similar to the one your have posted. I like the 'fat man' analogy. I thought it was pretty funny being a skinny guy.

ger21,
I was wondering how I would have to wait until I got your sage advice. I will model the spindle and mounting plates and see if I can free up some room. Great idea.

One, you're still going to get some flex, as the spindle will still be mounted to the end of a lever. It's just a shorter lever.

I realize that. My hope is that I will have thick enough material that the flex is negligent. That's the goal anyway...

And two, if your spindle can reach the bottom of your table, I doubt it will clear the bottom of the gantry when fully retracted.

In my previous design iteration the spindle did clear the bottom of the gantry when retracted. I will have to check it again for the new design though. Thanks for the reminder.

[ger21]

In my previous design iteration the spindle did clear the bottom of the gantry when retracted. I will have to check it again for the new design though. Thanks for the reminder.

Not just the spindle, but the tip of the tool as well need to clear any workpiece.

[amishx64]

Not just the spindle, but the tip of the tool as well need to clear any workpiece.

A tool with what approx length? An inch or two? I don't have many end mills or other tools yet so I genuinely don't know. What would you recommend?

[ger21]

Ideally, you want to check with both the longest, and shortest tools you intend to use.
If you use a tool that extends 3" from the collet, then you want it to clear the bottom of the gantry.
If you use a tool that extends 1" from the collet, you want to be sure it will reach your table.